GPS-Denied ISAC Vehicle Localization Based on mmWave Radar and Identification

Millimeter wave (mmWave) radar has become a widely adopted technology in vehicles and advanced driver assistance systems (ADAS). Meanwhile, as wireless communication progresses to higher frequencies, the role of the mmWave band has expanded, now supporting both sensing and 5G communication, which has given rise to integrated sensing and communication (ISAC). In this paper, we propose a novel mmWave ISAC vehicle localization system, which innovatively integrates emerging mmWave identification (MMID) technology into conventional automotive mmWave radar, enabling automotive radar to interact with roadside MMID tags, thereby achieving lane-level vehicle localization independent of the global positioning system (GPS). Unlike existing RFID-based vehicle localization solutions, the proposed solution is more practical for real-world deployment, as it eliminates the need to install additional large RFID antennas on vehicles. To achieve this, we first analyze the backscatter modulation characteristics of MMID tags and propose a novel frequency modulation strategy that lays the foundation for distinguishing signals from different tags within radar echoes that contain various tags and other objects. Subsequently, based on the distance, relative velocity, and azimuth of the tags, we perform static parameter estimation of the vehicle using the least squares (LS) algorithm. Finally, we construct a vehicle motion model and introduce a novel mutation particle filter (MPF) algorithm to estimate the dynamic motion state of the vehicle, ultimately achieving precise vehicle position tracking. The proposed system offers a practical solution for GPS-denied vehicle localization, aligning with the future vision of 6G-enabled intelligent transportation systems (ITS) and IoT-driven smart cities.